[src/trunk]: src/lib/libm/src There is some very odd code in s_exp2.c that on...

[dsl <dsl%NetBSD.org@localhost>]

details:   https://anonhg.NetBSD.org/src/rev/9f258243789d
branches:  trunk
changeset: 327704:9f258243789d
user:      dsl <dsl%NetBSD.org@localhost>
date:      Fri Mar 14 22:21:31 2014 +0000

description:
There is some very odd code in s_exp2.c that only works if 'double'
   values are rounded to 53 bit mantissa in teh x87 registers.
This hasn't been true since Nov 11 2013.
Forcing a store-load for 'double' in STRICT_ASSIGN() seems to fix things
  (at least enough for the few random test cases).
I suspect the code could be changed so that the number of mantissa bits
  didn't matter.

diffstat:

 lib/libm/src/math_private.h |  4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diffs (21 lines):

diff -r a3aeb90e0b00 -r 9f258243789d lib/libm/src/math_private.h
--- a/lib/libm/src/math_private.h       Fri Mar 14 21:59:41 2014 +0000
+++ b/lib/libm/src/math_private.h       Fri Mar 14 22:21:31 2014 +0000
@@ -11,7 +11,7 @@
 
 /*
  * from: @(#)fdlibm.h 5.1 93/09/24
- * $NetBSD: math_private.h,v 1.20 2013/11/19 19:24:34 joerg Exp $
+ * $NetBSD: math_private.h,v 1.21 2014/03/14 22:21:31 dsl Exp $
  */
 
 #ifndef _MATH_PRIVATE_H_
@@ -185,7 +185,7 @@
 #define        STRICT_ASSIGN(type, lval, rval) do {    \
        volatile type __lval;                   \
                                                \
-       if (sizeof(type) >= sizeof(double))     \
+       if (sizeof(type) >= sizeof(long double))        \
                (lval) = (rval);                \
        else {                                  \
                __lval = (rval);                \